Efficient Model for Probabilistic Web Resources Under Uncertainty

Abstract

Individuals, organizations, and connected objects produce and publish vast data on the web. These data are then combined into mashups to produce new valuable data. However, combining data from different sources may lead to data uncertainty because these sources contain heterogeneous, contradictory, or incomplete information. In expert and intelligent systems, the word “uncertainty” is related to working with inaccurate data, imprecise and incomplete information, and unreliability of results that can lead to irrational decisions. An expert and intelligent system can consider all the uncertain information, and manage web services and web resources based on the best possible answer rather than on the quality of the exact answer to manage such problems. This paper proposes a new model to define, compute, and interpret uncertain web resources in the context of classical hypertext navigation and evaluation of data queries. The experimental study as well as the analysis of the results, which we carried out on different corpora, confirmed the usefulness of our approach. It provides the effect of the treatment of uncertainty on the execution time and the importance of calculating the confidence level of the response returned to the user. The analysis of the IT execution time shows that the time required for the composition of the services by our approach is negligible, as compared to that of the other approach studied. The impact of the variation in the number of nodes of type “mux” and type “ind” were also evaluated. Our algorithm checks all possibilities in polynomial time and can adapt to many possibilities of multiplexing values.